This is the peer reviewed version of the following article: Angew. Chem. Int. Ed. 2016, 55, 2764-2767, which has been
published in the final form at http://onlinelibrary.wiley.com/doi/10.1002/anie.201510540/full. This article may be used for noncommercial purposes in accordance with Wiley Terms and Conditions for Self-Archiving :
(https://authorservices.wiley.com/author-resources/Journal-Authors/licensing-open-access/open-access/self-archiving.html)

Cooperative Reductive Elimination: The Missing Piece in the
Oxidative Coupling Mechanistic Puzzle
Ignacio Funes-Ardoiz,[a] and Feliu Maseras*[a,b]
Abstract:
The
reaction
between
benzoic
acid
and
methylphenylacetylene to form an isocoumarin is catalyzed by
Cp*Rh(OAc)2 in the presence of Cu(OAc)2(H2O) as an oxidant and a
leading example of oxidative-coupling reactions. Its mechanism was
elucidated by DFT calculations with the B97D functional. The
conventional mechanism, with separate reductive-elimination and
reoxidation steps, was found to yield a naphthalene derivative as the
major product by CO2 extrusion, contradicting experimental
observations. The experimental result was reproduced by an
alternative mechanism with a lower barrier: In this case, the copper
acetate oxidant plays a key role in the reductive-elimination step,
which takes place through a transition state containing both rhodium
and copper centers. This cooperative reductive-elimination step
would not be accessible with a generic oxidant, which, again, is in
agreement with available experimental data.

Direct catalytic C-H activation is an appealing option for the
synthesis of complex organic molecules. 1 In contrast with the
widely applied cross-coupling,2,3 it does not require the
placement of a halide or any other leaving group in the position
to be activated, and it improves atom economy by reducing the
formation of waste products. When combined with the facile
activation of an acidic O-H or N-H bond in the same molecule,
direct C-H activation opens the way to annulative coupling
processes with great potential.4 However, the common need for
an external oxidant has hindered the further development of
such processes. If the two hydrogen atoms depart as protons,
as is normally the case, the catalytic system will gain two
electrons, which then have to be removed prior to a new
catalytic cycle. Oxidative-coupling processes catalyzed by
transition-metal complexes based on palladium,4a,5 rhodium,6
ruthenium4b,7 and copper8 have been reported, but the issue of
the oxidant remains challenging.
Such reactions only seems to be efficient only with certain
combinations of catalyst and oxidant. For instance, in reactions
with the Rh(III)/Rh(I) redox pair as the catalyst, the oxidant of
choice is usually the CuII species Cu(OAc)2·H2O, which is
reduced to a CuI complex during the reaction.9-11 The copper salt
has been postulated to play an active role,11 but detailed
mechanistic evidence is scarce. Jones and co-workers showed

[a]

[b]

M.Sc. I. Funes-Ardoiz, Prof. F. Maseras
Institute of Chemicas Research of Catalonia (ICIQ), The Barcelona
Institute of Science and Technology.
Avgda. Països Catalans, 16, Tarragona (43007), Catalonia, Spain
E-mail: fmaseras@iciq.es
Prof. F. Maseras
Department de Química
Universitat Autònoma de Barcelona
Bellaterra (08193) Catalonia, Spain.

Supporting information for this article is given via a link at the end of
the document.

by UV-vis monitoring that a Rh diacetate species is formed in
the presence of high acetate concentrations.12 The presence of
additional acetate, introduced as NaOAc, was also shown to be
critical for the isolation of intermediates in the oxidative coupling
of isoquinolones with alkynes.13 The oxidant role of the copper
center has not been examined in detail in either of these
experimental studies or in the computational analyses14 that
have been carried out on related processes with RhIII/CuII
systems. Herein, we computationally examine the specific role of
CuII in the mechanism of this process.
For our study, we chose one of the leading examples of this
type of chemistry, an oxidative coupling reported by Miura,
Satoh, and Ueura, which is shown in Scheme 1. 9 We were
particularly interested in the effect of the oxidant on the
selectivity of the reaction.9b When copper diacetate is used as
oxidant, the isocoumarin 3 is formed in high yield. In contrast,
when other oxidants, such as Ag(OAc, are used, CO2 extrusion
takes place, resulting in naphthalene derivative 4 as the major
product.

Scheme 1. Reported oxidative coupling between benzoic acid and alkyne.9

We studied the key steps in this process through geometry
optimizations in ortho-xylene solution with the B97D functional.
The only simplification from the real system consisted in
replacing Cp* by Cp. A functional calibration showed that the
results were similar to those obtained with MO6-D3, B3LYP,D3,
BP86-D3, and B97x-D (See the Supporting Information for full
computational details and information on functional and basis
set calibration).
The starting point for our mechanistic study was the proposal
put forward by Miura, Satoh, and Ueura, which is summarized in
Scheme 2. The reaction is initiated by protonation and
displacement of one of the acetate ligands in the catalyst by
benzoic acid. The second acetate participates in a concerted
metallation deprotonation (CMD) step, which leads to cleavage
of the ortho C-H bond.15 Insertion of an alkyne into the Rh-C
bond then leads to intermediate 5, which contains a sevenmembered cyclometalated ring. All of these steps leading up to
intermediate 5 have been well characterized for related systems
as having low barriers. This hypothesis was confirmed for this
system by calculations that are shown in the Supporting
Information.

examine the possibility of the CuII reducing agent interacting
directly with 5. This was not trivial, as Cu(OAc)2.H2O does not
exist primarily as a monometallic complex in solution. It is
mainly in a dimetallic state, with the additional complication that
the acetate ligand may exchange with other ligands, such as
chloride (from the Cp*RhCl2 catalyst precursor), that are also
available in the reaction mixture. We found that both
[Cu(OAc)2(H2O)]2 and [CuCl(OAc)(H2O)]2 form stable adducts
with intermediate 5. The second of these adducts, with the
chlorinated copper species, leads to the productive pathway for
reductive elimination indicated in Scheme 3, with the highest
free energy point in TS6-7 only 0.6 kcal/mol relative to 5. This is
much lower than the energy of the transition state of the
alternative CO2 extrusion pathway.

Scheme 2. Initial experimental proposal for the mechanism.

Figure 1 shows the structure of the key transition state TS67 in this favored pathway. In this transition state, as well as in
the related intermediates 6 and 7 (see the Supporting
Infromation), there is a direct interaction between one of the
carboxylate oxygen atoms and one of the copper centers, which

Intermediate 5 can evolve through one of the two alternative
paths depicted in Scheme 2 to yield one of the two possible
products, 3 or 4, and to regenerate the catalyst. We
computationally studied the reactivity of 5 and found, to our
surprise, that the path leading to CO2 extrusion is clearly favored,
as shown in the free energy profiles in Scheme 3. The direct
reductive elimination pathway through Ts5-8, with geometrical
features usually associated with these processes,16 has a free
energy barrier of 13.0 kcal/mol relative to 5. The value is modest,
but clearly higher than that for CO2 extrusion through Ts5-9,
which is 8.2 kcal/mol relative to 5. We analyzed possible
alternative paths that involve the coordination of acetic acid in
the reaction medium, but the barrier for reductive elimination
was even higher (26.5 kcal/mol, see the Supporting Information,
Figure S3).

Figure 1. Key transition state (TS 6-7) of the cooperative reductive elimination
pathway with spin densities and bon lengths (in Å).

Scheme 3. Free energy profile of the selectivity determining step in the
reaction. All energies in kcal/mol referred to the intermediate 5.

As intermediate 5 prefers to evolve towards CO2 extrusion,
which contradicts the experimental observation, we decided to

hinders CO2 extrusion. The structure clearly corresponds to C-O
formation, as indicated by the computed normal mode
associated to the imaginary frequency and the C-O bond
distance of 1.89 Å. The bridging position of the chlorine atom
between Cu and Rh, with M-Cl distances of 2.47 and 2.48 Å,
respectively, is also noteworthy. This bridge stabilizes the
adduct and favors the electron transfer between the metal
centers.

Figure 1 also reports the spin densities, which provide
information on the underlying features of this step. Two unpaired
electrons are present because each initial CuII center contains
one unpaired electron, and our calculations indicate that the
dimer is most stable in a triplet spin state, as expected. The
triplet lies 0.1 kcal/mol in energy below the open-shell singlet
and 6.2 kcal/mol below the closed-shell singlet. The ground
state for species 6, TS6-7, and 7 is a triplet. The open-shell
singlet state for the key species TS6-7 is 1.1 kcal/mol above the
triplet state with a qualitatively similar spin distribution and thus
not discussed here. Concerning the triplet ground state, Figure 1
indicates the presence of significant spin density on rhodium
(0.13) and one of the carbon atoms coming from the alkyne
(0.28). Remarkably, the spin density on the two copper centers
is only mildly higher (0.28, 0.36). The presence of spin density
on the rhodium moiety of the adduct strongly suggests that the
RhIII/RhI description is not appropriate for the process taking
place, as both of these oxidation states are associated with
closed-shell distributions. Instead, we favour a scenario where
RhIII is partially reduced to RhII, and the other electron is
transferred directly to the dicopper unit, with one Cu II and one
CuI center. A similar scenario would be ontained with the openshell singlet spin state, but at a slightly higher energy (see the
Supporting Information). This process therefore is an unusual
reductive elimination, as summarized in Scheme 4. Only one of
the two electrons from the anionic organic ligands goes to the
adjacent metal center, while the other one is transferred directly
to the final oxidant. We thus refer to this process as cooperative
reductive elimination (CRE).

It follows from the description above that this mechanism
can only operate with very specific reducing agents, which
explains why oxidative coupling did not take place in the
systems under study when other oxidizing agents were used
instead of copper diacetate.9 The fact that some of the acetate
ligands are replaced by chloride in the active transition state also
agrees with the observed influence of acetate concentration on
the nature of species in solution.12,13 The participation of more
than one metal center in the reduction is reminiscent of the
bimetallic reductive elimination observed in dinuclear PdIII
complexes,17 but with substantial differences. The role of the
additive is also different from that computationally determined for
AgOAc and CsF in PdII C-H activation chemistry.18
In conclusion, we have described a new cooperative
reductive-elimination mechanism that satisfactorily explains the
specific role of the copper diacetate oxidant in the rhodiumcatalyzed oxidative coupling of benzoic acid and alkynes. The
oxidating agent participates directly in the reductive elimination
process by taking one electron already in the reductiveelimination transition state itself. In this way, the rhodium I
oxidation state is never reached, which constitutes a clear
deviation from the commonly observed reductive-elimination
pattern at transition-metal centers. We consider that this
cooperative reductive-elimination mechanism may be operating
in other systems where external oxidazing agents are involved.
Ongoing research in our group will aim to clarify the eventual
generality of this mechanism.

Acknowledgements
Financial support from the ICIQ Foundation, MINECO (project
CTQ2014-57661-R, and Severo Ochoa Excellence Accreditation
2014-2018 SEV-2013-0319) is gratefully acknowledged. I.F-A.
acknowledges a fellowship from the ICIQ-Severo Ochoa
program (Ref. SVP-2014-068662). We also thank Dr. Max
García-Melchor and Dr. Nuno A.G. Bandeira for helpful
discussions.
Keywords: Oxidative coupling • Density functional calculations
• Reductive elimination • C-C coupling • Reaction mechanisms
Scheme 4. Contrast between the conventional mechanism (reductive
elimination plus reoxidation, above) and the cooperative reductive elimination
mechanism (below).

The spin densities in 6 and 7 (figure S4) confirm the validity
of this scenario. The total spin density in the two Cu centers
decreases from 0.81 to 0.57 when going from 6 to 7, while the
density at Rh increases from 0.14 to 0.59. One of the semioccupied molecular orbitals in TS 6-7 is also mostly located in
the Rh moiety of the system (Figure S5), corroborating the CRE
proposal. After the reductive elimination, the trimetallic complex
7 reorganizes; the organic product 3 is released (Scheme S6)
and the Rh center then transfers one electron, with assistance
from the Cl bridge atoms, to regenerate the initial catalyst
[CpRhCl2] and the [Cu(OAc)(H2O)]2 waste product.

[1]

[2]
[3]
[4]
[5]
[6]

[7]

a) D. Alberico, M. E. Scott, M. Lautens, Chem. Rev. 2007, 107, 174238; b) L. Ackermann, R. Vicente, A. R. Kapdi, Angew. Chem. Int. Ed.
2009, 48, 9792-9826; c) J. Wencel-Delord, F. Glorius, Nature Chem.
2013, 5, 369-375.
Metal-Catalyzed Cross-Coupling Reactions, 2nd Ed. (Eds.: A. de Meijere,
F. Diedrich), Wiley-VCH, Weinheim, 2004.
M. Garcia-Melchor, A. A. C. Braga, A. Lledos, G. Ujaque, F. Maseras,
Acc. Chem. Res. 2013, 46, 2626-2634.
a) T. Satoh, M. Miura, Chem. Eur. J. 2010, 16, 11212-11222; b) L.
Ackermann, Acc. Chem. Res. 2014, 47, 281-295.
a) T. W. Lyons, M. S. Sanford, Chem. Rev. 2010, 110, 1147-1169; b) D.
C. Powers, T. Ritter, Acc. Chem. Res. 2012, 45, 840-850.
a) D. A. Colby, A. S. Tsai, R. G. Bergman, J. A. Ellman, Acc. Chem.
Res. 2012, 45, 814-825; b) B. Ye, N. Cramer, Acc. Chem. Res. 2015,
48, 1308-1318;
P. B. Arockiam, C. Bruneau, P. H. Dixneuf, Chem. Rev. 2012, 112,
5879-5918.

[8]
[9]
[10]

[11]

[12]
[13]

X. –X. Guo, D. –W. Gu, Z. Wu, W. Zhang, Chem. Rev. 2015, 115,
1622-1651.
a) K. Ueura, T. Satoh, M. Miura, Org. Lett. 2007, 9, 1407-1409; b) K.
Ueura, T. Satoh, M. Miura, J. Org. Chem. 2007, 72, 5362-5367;
a) D. R. Stuart, M. Bertrand-Laperle, K. M. N. Burgess, K. Fagnou, J.
Am. Chem. Soc. 2008, 130, 16474-16475; b) T. K. Hyster, T. Rovis, J.
Am. Chem. Soc. 2010, 132, 10565-10569; c) L. Huang, Q. Wang, J. Qi,
X. Wu, K. Huang, H. Jiang, Chem. Sci. 2013, 4, 2665-2669; d) J. Shi, Y.
Yan, L. Qiu, H. E. Xu, W. Yi, Chem. Commun. 2014, 50, 6483-6486; e)
J. Zheng, S. –B. Wang, C. Zheng, S. L. You, J. Am. Chem. Soc. 2015,
134, 4880-4883 ; f) R. B. Dateer, S. Chang, J. Am. Chem. Soc. 2015,
137, 4908-4911; g) A. M. Martínez, J. Echavarren, I. Alonso, N.
Rodriguez, R. Gómez-Arrayás, J. C. Carretero, Chem. Sci. 2015, 6,
5802-5814.
a) L. Li, W. W. Brennessel, W. D. Jones, J. Am. Chem. Soc. 2008, 130,
12414-12419; b) F. W. Patureau, J. Wencel-Delord, F. Glorius,
Aldrichimica Acta 2012, 2, 31-41.
L. Li, W. W. Brennessel, W. D. Jones, Organometallics 2009, 28, 34923500;.
N. Wang, B. Li, H. Song, S. Xu, B. Wang, Chem. Eur. J. 2013, 19, 358364.

[14]

[15]

[16]

[17]
[18]

a) L. Xu, Q. Zhu, G. Huang, B. Cheng, Y. Xia, J. Org. Chem. 2012, 77,
3017-3024; b) N. Quiñones, A. Seoane, R. García-Fandiño, J. L.
Mascareñas, M. Gulías, Chem. Sci. 2013, 4, 2874-2879; c) D. L.
Davies, C. E. Ellul, S. A. Macgregor, C. L. McMullin, J. Am. Chem. Soc.
2015, 137, 9659-9669; d) J. Jiang, R. Ramozzi, K. Morokuma, Chem.
Eur. J. 2015, 21, 11158-11164.
a) D. L. Davies, S. M. A. Donald, O. Al-Duaij, S. A. Macgregor, M.
Pölleth, J. Am. Chem. Soc. 2006, 128, 4210-4211. b) D. GarciaCuadrado, P. de Mendoza, A. A. C. Braga, F. Maseras, A. M.
Echavarren J. Am. Chem. Soc. 2007, 129, 6880-6886; c) S. I. Gorelsky,
D. Lapointe, K. Fagnou, J. Am. Chem. Soc. 2008, 130, 10848-10849.
a) V. P. Ananikov, D. G. Musaev, K. Morokuma, J. Am. Chem. Soc.
2002, 124, 2839-2852; b) M. Perez-Rodriguez, A. A. C. Braga, M.
Garcia-Melchor, M. H. Perez-Temprano, J. A. Casares, G. Ujaque, A. R.
de Lera, R. Alvarez, F. Maseras, P. Espinet, J. Am. Chem. Soc. 2009,
131, 3650-3657.
D. C. Powers, D. Benitez, E. Tkatchouk, W. A. Goddard III, T. J. Ritter,
J. Am. Chem. Soc. 2010, 132, 14092-14103.
M. Anand, R. B. Sunoj, H. F. Schaefer, J. Am. Chem. Soc. 2014, 136,
5535-5538.

Entry for the Table of Contents

COMMUNICATION
Joint effort. The reaction between
benzoic acid and methylphenylacetylene to form an isocoumarin is
catalyzed by Cp*Rh(OAc)2 in the
presence of Cu(OAc)2(H2O) as the
oxidant.
Its
mechanism
was
elucidated by DFT calculations with
the B97D functional, which showed
that
the
overall
transformation
proceeds by cooperative reductive
elimination with a transition state
containing both rhodium and copper
centers.

Author(s), Corresponding Author(s)*
Page No. – Page No.
Cooperative Reductive Elimination:
The Missing Piece in the Oxidative
Coupling Mechanistic Puzzle